The parasitic protozoan Trypanosoma brucei is the causative agent of African trypanosomiasis which causes sleeping sickness in humans and Nagana in farm animals. The World Health Organization projects that the current disease burden of human African trypanosomiasis is about two million Disability Adjusted Life Years (with 300,000 new cases each year). Coupled with the profound effect on the economy caused by Nagana, T. brucei is one of the leading impediments to development in much of Africa. While related parasites utilize protected niches after they enter their mammalian host, T. brucei thrives in the mammalian bloodstream, where it must survive the onslaught of the humoral immune response. The parasite does this by virtue of its dense surface coat, consisting of ~5 x 106 variable surface glycoprotein (VSG) homodimers. While this VSG coat is exposed to and recognized by antibodies, T. brucei periodically replaces it (in a process called VSG switching) enabling a subpopulation of parasites to survive by evading complement-mediated lysis. Little is known about VSG switching at the molecular level. Work over the last 25 years has determined (a) that the majority of switching occurs by duplicative gene conversion and (b) that there is semi-predictable hierarchy with regard to the types of VSGs that appear early vs. late after infection. We have recently established a robust experimental system, which we have used to successfully reproduce many of the features of in vivo infection. Here, we propose to use this system toward two specific aims: 1) Determine the mechanism that initiates VSG switching. There is a long-standing hypothesis that VSG switching is initiated by an endonuclease. Alternatively, switching could be a result of spontaneous but frequent DNA breaks that can arise from endogenous processes peculiar to chromosome ends (which is where all expressed VSGs are located). We propose experiments to evaluate both hypotheses. 2) Understand the rules that govern choice of donor VSG, leading to the observed hierarchy of switching. In the early stages of VSG switching, the new (donor) VSGs arising in the course of infection are not random. For example, cells expressing VSG221 tend to switch to VSG224, both in vivo (Robinson et al., 1999) and in vitro (our preliminary data). We will evaluate two mechanisms that could explain this consistent finding: (a) initial preference could be due to sequence similarity in conserved elements (such as the 70-bp repeat tracts) of expression sites, and (b) this preference could be the result of proximity of the two expression sites in the context of the nucleus. African trypanosomiasis is always fatal unless treated. The few therapeutic treatment options that do exist are expensive, have serious side effects and are increasingly inefficient as drug resistant T. brucei strains begin to emerge. The work we propose centers upon understanding and eventually disrupting the major known pathway of immune evasion by this parasite which is also the cause of pathogen persistence. As such, in the long term the research we propose could lead to a novel and effective way to protect against this neglected disease.